Ouranosinc / xscen / 19972609741

90%

push

github

Reduce warnings (#665)

<!-- Please ensure the PR fulfills the following requirements! -->
<!-- If this is your first PR, make sure to add your details to the
AUTHORS.rst! -->
### Pull Request Checklist:
- [ ] This PR addresses an already opened issue (for bug fixes /
features)
    - This PR fixes #xyz
- [x] (If applicable) Documentation has been added / updated (for bug
fixes / features).
- [x] (If applicable) Tests have been added.
- [x] This PR does not seem to break the templates.
- [x] CHANGELOG.rst has been updated (with summary of main changes).
- [x] Link to issue (:issue:`number`) and pull request (:pull:`number`)
has been added.

### What kind of change does this PR introduce?

* Solves a few warnings primarily coming from `xarray`.

### Does this PR introduce a breaking change?

* dtype encodings are now removed before saving, since it could create
inconsistencies. It can be accomplished through kwargs instead,
* For example, the `air` variable of the xarray tutorial dataset has a
`int64` in its encoding, when it's clearly a float and recognized as
such by xarray.

### Other information:

* I'm not sure what the right fix is for `pd.concat([dc.df for dc in
data_catalogs], ignore_index=True)`. The warning was that in the future,
NaNs would be taken into account when determining the dtype of a column.
In our case, this can happen if we open `simulation` and
`reconstruction` catalogs. I tried to put the right dtypes, but then I
had issues with subsequent tests, so I've left it as `object` mostly
everywhere.
* For the various `concat`/ `merge`, I've set them as the current
defaults.

12 of 12 new or added lines in 5 files covered. (100.0%)

4237 of 4691 relevant lines covered (90.32%)

3.61 hits per line